1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
|
---
layout: global
title: Building Spark with Maven
---
* This will become a table of contents (this text will be scraped).
{:toc}
Building Spark using Maven requires Maven 3.0.4 or newer and Java 1.6 or newer.
## Setting up Maven's Memory Usage ##
You'll need to configure Maven to use more memory than usual by setting `MAVEN_OPTS`. We recommend the following settings:
export MAVEN_OPTS="-Xmx2g -XX:MaxPermSize=512M -XX:ReservedCodeCacheSize=512m"
If you don't run this, you may see errors like the following:
[INFO] Compiling 203 Scala sources and 9 Java sources to /Users/me/Development/spark/core/target/scala-{{site.SCALA_BINARY_VERSION}}/classes...
[ERROR] PermGen space -> [Help 1]
[INFO] Compiling 203 Scala sources and 9 Java sources to /Users/me/Development/spark/core/target/scala-{{site.SCALA_BINARY_VERSION}}/classes...
[ERROR] Java heap space -> [Help 1]
You can fix this by setting the `MAVEN_OPTS` variable as discussed before.
*Note: For Java 1.8 and above this step is not required.*
## Specifying the Hadoop version ##
Because HDFS is not protocol-compatible across versions, if you want to read from HDFS, you'll need to build Spark against the specific HDFS version in your environment. You can do this through the "hadoop.version" property. If unset, Spark will build against Hadoop 1.0.4 by default. Note that certain build profiles are required for particular Hadoop versions:
<table class="table">
<thead>
<tr><th>Hadoop version</th><th>Profile required</th></tr>
</thead>
<tbody>
<tr><td>0.23.x</td><td>hadoop-0.23</td></tr>
<tr><td>1.x to 2.1.x</td><td>(none)</td></tr>
<tr><td>2.2.x</td><td>hadoop-2.2</td></tr>
<tr><td>2.3.x</td><td>hadoop-2.3</td></tr>
<tr><td>2.4.x</td><td>hadoop-2.4</td></tr>
</tbody>
</table>
For Apache Hadoop versions 1.x, Cloudera CDH "mr1" distributions, and other Hadoop versions without YARN, use:
# Apache Hadoop 1.2.1
$ mvn -Dhadoop.version=1.2.1 -DskipTests clean package
# Cloudera CDH 4.2.0 with MapReduce v1
$ mvn -Dhadoop.version=2.0.0-mr1-cdh4.2.0 -DskipTests clean package
# Apache Hadoop 0.23.x
$ mvn -Phadoop-0.23 -Dhadoop.version=0.23.7 -DskipTests clean package
For Apache Hadoop 2.x, 0.23.x, Cloudera CDH, and other Hadoop versions with YARN, you can enable the "yarn-alpha" or "yarn" profile and optionally set the "yarn.version" property if it is different from "hadoop.version". The additional build profile required depends on the YARN version:
<table class="table">
<thead>
<tr><th>YARN version</th><th>Profile required</th></tr>
</thead>
<tbody>
<tr><td>0.23.x to 2.1.x</td><td>yarn-alpha</td></tr>
<tr><td>2.2.x and later</td><td>yarn</td></tr>
</tbody>
</table>
Examples:
# Apache Hadoop 2.0.5-alpha
$ mvn -Pyarn-alpha -Dhadoop.version=2.0.5-alpha -DskipTests clean package
# Cloudera CDH 4.2.0
$ mvn -Pyarn-alpha -Dhadoop.version=2.0.0-cdh4.2.0 -DskipTests clean package
# Apache Hadoop 0.23.x
$ mvn -Pyarn-alpha -Phadoop-0.23 -Dhadoop.version=0.23.7 -DskipTests clean package
# Apache Hadoop 2.2.X
$ mvn -Pyarn -Phadoop-2.2 -Dhadoop.version=2.2.0 -DskipTests clean package
# Apache Hadoop 2.3.X
$ mvn -Pyarn -Phadoop-2.3 -Dhadoop.version=2.3.0 -DskipTests clean package
# Apache Hadoop 2.4.X
$ mvn -Pyarn -Phadoop-2.4 -Dhadoop.version=2.4.0 -DskipTests clean package
# Different versions of HDFS and YARN.
$ mvn -Pyarn-alpha -Phadoop-2.3 -Dhadoop.version=2.3.0 -Dyarn.version=0.23.7 -DskipTests clean package
## Spark Tests in Maven ##
Tests are run by default via the [ScalaTest Maven plugin](http://www.scalatest.org/user_guide/using_the_scalatest_maven_plugin). Some of the require Spark to be packaged first, so always run `mvn package` with `-DskipTests` the first time. You can then run the tests with `mvn -Dhadoop.version=... test`.
The ScalaTest plugin also supports running only a specific test suite as follows:
$ mvn -Dhadoop.version=... -Dsuites=org.apache.spark.repl.ReplSuite test
## Continuous Compilation ##
We use the scala-maven-plugin which supports incremental and continuous compilation. E.g.
$ mvn scala:cc
should run continuous compilation (i.e. wait for changes). However, this has not been tested extensively.
## Using With IntelliJ IDEA ##
This setup works fine in IntelliJ IDEA 11.1.4. After opening the project via the pom.xml file in the project root folder, you only need to activate either the hadoop1 or hadoop2 profile in the "Maven Properties" popout. We have not tried Eclipse/Scala IDE with this.
## Building Spark Debian Packages ##
The maven build includes support for building a Debian package containing the assembly 'fat-jar', PySpark, and the necessary scripts and configuration files. This can be created by specifying the following:
$ mvn -Pdeb -DskipTests clean package
The debian package can then be found under assembly/target. We added the short commit hash to the file name so that we can distinguish individual packages built for SNAPSHOT versions.
## Running java 8 test suites.
Running only java 8 tests and nothing else.
$ mvn install -DskipTests -Pjava8-tests
Java 8 tests are run when -Pjava8-tests profile is enabled, they will run in spite of -DskipTests.
For these tests to run your system must have a JDK 8 installation.
If you have JDK 8 installed but it is not the system default, you can set JAVA_HOME to point to JDK 8 before running the tests.
## Packaging without Hadoop dependencies for deployment on YARN ##
The assembly jar produced by "mvn package" will, by default, include all of Spark's dependencies, including Hadoop and some of its ecosystem projects. On YARN deployments, this causes multiple versions of these to appear on executor classpaths: the version packaged in the Spark assembly and the version on each node, included with yarn.application.classpath. The "hadoop-provided" profile builds the assembly without including Hadoop-ecosystem projects, like ZooKeeper and Hadoop itself.
|